Classifications

• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1602—Programme controls characterised by the control system, structure, architecture
  • B25J9/161—Hardware, e.g. neural networks, fuzzy logic, interfaces, processor
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J19/00—Accessories fitted to manipulators, e.g. for monitoring, for viewing; Safety devices combined with or specially adapted for use in connection with manipulators
  • B25J19/007—Means or methods for designing or fabricating manipulators
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/08—Programme-controlled manipulators characterised by modular constructions
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1602—Programme controls characterised by the control system, structure, architecture
  • B25J9/1605—Simulation of manipulator lay-out, design, modelling of manipulator
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1628—Programme controls characterised by the control loop
  • B25J9/1653—Programme controls characterised by the control loop parameters identification, estimation, stiffness, accuracy, error analysis
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
  • B25J9/1661—Programme controls characterised by programming, planning systems for manipulators characterised by task planning, object-oriented languages
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1679—Programme controls characterised by the tasks executed
  • B25J9/1692—Calibration of manipulator
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
  • G06Q10/00—Administration; Management
  • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
  • G06Q10/063—Operations research, analysis or management
  • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
  • G06Q10/06311—Scheduling, planning or task assignment for a person or group
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
  • G06Q10/00—Administration; Management
  • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
  • G06Q10/063—Operations research, analysis or management
  • G06Q10/0633—Workflow analysis
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1615—Programme controls characterised by special kind of manipulator, e.g. planar, scara, gantry, cantilever, space, closed chain, passive/active joints and tendon driven manipulators
  • B25J9/1617—Cellular, reconfigurable manipulator, e.g. cebot
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B23/00—Testing or monitoring of control systems or parts thereof
  • G05B23/02—Electric testing or monitoring
  • G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
  • G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
  • G05B23/0283—Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL]
• • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
  • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
  • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
  • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
  • Y02P90/80—Management or planning

Definitions

• the invention relates to the field of production systems, in particular to manufacturing systems comprising robots or manipulators.
• manipulators are assigned to a specific tasks and cannot be easily adapted in terms of degree of freedom of movement, geometry, or of mechanical / physical capabilities to perform other kinds of tasks. Due to costs pressure, robots or manipulators are broadly used in production and companies want to maximize their ROI when buying robots. On the other hand, the popularization of robots creates the issue of recycling them, and environmental questions need to be considered. Hence, there is a strong need to
• US 2004/0073468 Al discloses a fleet management system to manage a plurality of machines aimed towards preventive maintenance of the machines.
• the system comprises a data repository containing status and operating data collected by sensors which is then analysed to recommend a modification in the maintenance schedule.
• US 7143007 B2 discloses a method to monitor the lifecycle of a component of an equipment.
• the duty profile of a component is a means to estimate theoretical life of the component and actual life of the component is estimated by considering the operating information of the component.
• An adjusted theoretical useful life is computed and a scheduled replacement need is signalled when the amount of the theoretical useful life consumed is within a replacement range of the adjusted theoretical useful life.
• US 8533018 B2 discloses a system to manage construction machines having plurality of components. It comprises a schedule generation unit to generate a maintenance schedule, a judgment unit to decide on the schedule based on the availability data of components, and a correction unit to correct the schedule table if it is decided by the judgement unit.
• US 2014/0067108 discloses systems and methods for the dynamic control of task assignments in a fabrication process that employs a plurality of machines to fabricate a manufactured component.
• the solution provided includes tasks assignments to an available portion of a plurality of machines.
• a process variable reflecting the change of status of machines is monitored, and enables to define the portion of available machines dynamically, and to re-distribute the tasks accordingly.
• US 8428777 Bl discloses a method and system for distributing tasks among robotic system. It aims at optimizing execution time of a task and provides potential solutions of an alternative repartition of tasks within a group of available robotic devices based on their ranking. The ranking specifies an amount of usage of the devices over time.
• WO2017/149500 Al discloses a method for maintenance of a machine such as an ATM. Maintenance personnel is informed about maintenance operations to be performed. Historical data about e.g. type of maintenance, parts serviced, time of last maintenance can be stored in a central server. Parts can be identified by product ID.
• US 9'465'384 Bl describes a reconfigurable robotic workcell. It can be programmed by specifying tasks in terms of subtasks such as picking, moving, placing an object or applying a force. Certain parameters of subtasks can be left undefined in an offline programming phase, and be defined in a later programming phase, by interaction with the real robot in the workcell.
• US 2016/0176043 Al discloses selection of a manipulator from a group of manipulators, for performing a particular task. Selection is guided by stored data on success or failure, and time and power spent, for performing the task be each of the manipulators. The system as a whole can be considered "modular" in that several manipulators are available for use. Each manipulator by itself is however not described as being modular.
• US 2015/0239127 Al shows action planning for a robot, using a simulation module and graphic visualization.
• a robot environment can use sensors to determine a status of the robot and an object located in the workspace. The status is used in the simulation model, assisting a user in planning, or as input for an action planning module. Markers identifying objects or locations related to an action are superimposed on a real or simulated view of the workspace and parts.
• WO 2016/014774 Al describes cloud based robotic planning system that takes as input sensor values, a description of a task and a robot's configuration, and outputs instructions to the robot for carrying out the task. This is done on the basis of experiences made with another robot system, and simulations of the task. Simulation can take place in the cloud and/or as on-board simulation in the robot.
• robot system encompasses a production system and in particular a manufacturing system in which parts are manipulated and processed, e.g. by tooling and assembling.
• a system comprises robots or manipulators for handling the parts, and can comprise dedicated other machines such as machine tools, painting or welding machines etc.
• a robotic system can be a single manipulator, in particular a modular manipulator or robot, or a production cell, or a production plant, or even a group of geographically separated production plants.
• the method comprising operating a computer-based Inventory
• the Inventory is configured to operate as a part of or in co-operation with a computer-based system for controlling the operation of a robotic system
• the robotic system comprising one or more Hardware Modules to perform a task
• the Inventory comprising a plurality of Hardware Module Descriptions, each Hardware Module Description comprising
• a Hardware Module constitutes a smart pluggable module.
• a module being pluggable means that it can be connected both on a hardware level and on a software or communication level by means of a standardised interface.
• a module being smart means that it comprises a computing unit with data storage and data processing elements that allow the Hardware Module to, e.g. perform data processing, and with communication elements for communicating with other Hardware Modules.
• the computing unit can be implemented by means of a variety of hardware entities, from an embedded controller over a controlling computer to cloud based processing units.
• a Hardware Module can be designed to transmit information about its internal state, history of operation etc. to the Inventory.
• Hardware Modules can be configured to receive software updates and/or configuration updates in order to maintain compatibility with other Hardware Modules.
• Hardware Modules can be configured to receive software upgrades for adding new functionalities.
• This can be data processing software for a sensor, e.g. image analysis software for a camera, or a new motor control algorithm for a manipulator module.
• a Hardware Module can be any hardware Module.
• CRC central computation and command unit
• a robotic arm or manipulator subsystem with its own control unit, typically a controlling computer that is separate from the arm itself, acting as the computing unit, and an interface unit (also referred to as "compatibiliser unit” or translator) which presents a standardised interface to other Hardware Modules.
• This interface makes the manipulator subsystem appear within the robotic system in essentially the same way as a modular manipulator assembly or system (see below) appears. In this way, a robot from a third party, which itself is not modular, can be integrated into the overall robotic system as one module.
• a sensor module comprising at least one sensor, and an embedded controller as the computing unit, for processing raw sensor data acquired by the sensor and communicating with other Hardware Modules, in particular for transmitting processed sensor data to other Hardware Modules.
• a Hardware Module that comprises a joint and actuator for e.g. rotary or translational movement is called an active module.
• a Hardware Module that has a fixed geometry or fixed geometric configuration is called passive module.
• a sensor module as a rule is a passive module.
• the Hardware Module Descriptions can be stored in the Hardware Modules themselves, and/or in a database together with Hardware Module Descriptions from other Hardware Modules, from which the descriptions can be retrieved given the identifier of the Hardware Module
• the unique identifier can be an IP address or MAC address.
• the database can be a centralised database or a distributed database, in particular a cloud database.
• the Inventory implements a library that records for each module its history and its specifications etc, which allows planning and optimising maintenance. Furthermore, information regarding the same type of Hardware Module can be exchanged for further optimisation of maintenance and for exchanging solutions when malfunctions or other problems occur.
• the Inventory can be linked to an ERP and/or maintenance planning system for planning the deployment and/or maintenance of the Hardware Modules.
• the Inventory comprises process definitions, wherein each process definition is associated with a task and specifies one or more actions and/or subtasks that when executed accomplish the task, and for each action, Hardware Modules and/or Software Modules required for executing the action. This allows to represent process knowledge in the Inventory, and provide this information to methods for designing a robotic system using the modules and/or to methods for planning operation of a robotic system.
• the Inventory comprises descriptions of robotic assemblies, wherein a robotic assembly is a configuration of manipulator modules and the description of a robotic assembly comprises a description of the configuration.
• the description of the configuration can comprise a description of the Hardware Modules or types of Hardware Modules, the geometric relation in which they are assembled, and any Software Modules that are part of the robotic assembly.
• the robotic assembly can be linked to all the historical data of its modules. This allows the system not only to leam about the history of each Hardware Module and type of Hardware Module and draw conclusions, e.g. for planning maintenance, but also about the history of each configuration. For example, if robotic assemblies of a particular configuration repeatedly experiences failures in the same Hardware Module, this is a property of this configuration, and is relevant for all assemblies having the same configuration, e.g. for predictive maintenance.
• the description of physical characteristics of the Hardware Module comprises one or more of mechanical, electrical and component parameters such as:
• the physical characteristics usually are determined by the physical construction of the Hardware Module and its components. They can remain fixed over time, or they can change. A change can be detected by sensors in the Hardware Module itself, or by interaction of the Hardware Module with other Hardware Modules - which can be manipulator modules and/or sensor modules - in particular by performing calibration routines. For each parameter, a current value can be stored, and optionally also historic data with previous values. This represents changes of the parameter over time.
• the description of a current status of the Hardware Module comprises one or more of:
• the current status represents the status data, therefore status data comprises, e.g. data on the internal state, the other Hardware Modules that the Hardware Module is physically connected to, etc.
• the internal state can be a temperature inside the Hardware Module, the position of a joint that is part of the Hardware Module, which in this case can be a manipulator module, etc.
• the software modules associated with the Hardware Module can obviously be software that is executed on the Hardware Module, but they can also be "related" in that they process data acquired by the Hardware Module, e.g. image processing software for a hardware camera, or in that they determine data that is used in operation of the Hardware Module, e.g. calibration software that computes calibration parameters (from calibration data determined by the same Hardware Module, or from data obtained by other Hardware Modules, in particular, sensors).
• the operating data and the historical data representing usage of the Hardware Module comprises one or more of:
• the historical data can be derived from the operating data, and data can comprise data on when and how long a Hardware Module or components of the Hardware Module were in operation, and parameters used in their operation.
• operating data can state when device, e.g. a motor was switched on, what power it operated at, and when it was switched off.
• Operating data can comprise values of physical properties, in particular mechanical or electrical properties. Mechanical properties can be forces, torques, speeds, paths travelled, etc. Electrical properties can be currents, voltages, power, etc.
• Operating data can be determined by the embedded controller - for parameters that are controlled by the embedded controller itself - or from sensor data obtained from the at least one sensor.
• Historical data can also comprise entries that are in part or wholly generated by humans, such as how a particular malfunction was eliminated or a log of maintenance actions with the number, date, frequency of maintenance actions, and with individual steps taken during maintenance. Information about malfunctions can be generated automatically. If a Hardware Module is exchanged because of repairs or maintenance, it is replaced by a similar Hardware Module with a different identity, which is recorded by the system, either automatically or manually. If a Hardware Module is repaired but remains in place, this is recorded in its history. In embodiments, each Hardware Module monitors and provides its characteristics regarding lifetime of replaceable components, e.g. to the central computation and command unit (CC). It also maintains and communicates to the CCC the number of remaining operations which can be performed before the next component failure is expected. The CCC collects this information for each Hardware Module, and is able either to plan for maintenance before failure's occurrence, or to plan the replacement of relevant Hardware Modules in order to limit downtime.
• CCC central computation and command unit
• the CCC can obtain information about when a component failure is expected from the inventory. Such information can be detemiined by analysing historical data of components of the same type.
• a software module description can comprise a definition of an API (application programming interface), types or identifiers of Hardware Modules that the Software Module can interact with. Such interaction can be e.g. by the software being executed on a processor of the Hardware Module, or by being executed on another processor, but generating control commands that are input to the Hardware Module.
• API application programming interface
• the Hardware Module Description comprises a description of a communication interface of the Hardware Module.
• This description allows to establish communication with the Hardware Module. It can comprise an address under which the Hardware Module can be accessed. It can comprise an interface definition or API that defines how to access the functions provided by the Hardware Module.
• the method for operating a computer-based Inventory comprises the steps for designing a robotic system:
• Planning operation of the robotic system can be based on target tasks that are part of the performance requirements. This allows operation of the system to be flexible and to work around failures.
• the Hardware Module Descriptions can be Hardware Module Type Descriptions that describe Hardware Modules in a generic manner, i.e. with a description that is valid for all Hardware Modules of that type. This allows to perform the designing with such types, and to choose concrete Hardware Modules at a later stage.
• a concrete Hardware Module is an instance of a (generic) Hardware Module Type. That is, the concrete Hardware Module has its own unique identity, and a Hardware Module Description can be associated with this identity.
• one or more or all of the Hardware Module Descriptions can relate to concrete Hardware Modules.
• these concrete Hardware Modules are considered, in some cases compared to one another, and chosen.
• the set of Hardware Modules can be an existing robotic or production system that already exists in the physical configuration required for satisfying the performance requirements.
• the Inventory thus also serves as a database describing operational systems and not only individual, unrelated Hardware Modules.
• the Hardware Modules can provide for a variety of options. This allows to optimise the robotic system depending on what are the exact needs, i.e. performance requirements are. Modules can vary according the action performed in terms of size (S, M, L), in terms of mechanical strength, in terms of materials, in terms of memory size, etc.
• optimised robot for a particular task can be built up using, e.g. long life Hardware Modules with high strength for one task, medium quality modules for other, less demanding tasks. Overall, an optimised robotic system can give adequate performance for the lowest price.
• the performance requirements can comprise a 3D space to be reached by the robotic system, a maximum load to be manipulated by the robotic system, a maximum speed and accuracy for the movement of the load, etc.
• the step of automatically determining the set of Hardware Modules comprises the further steps of
• a process definition specifies actions that perform a task.
• a (target) task is an input to a design or planning method.
• a task is what the system is to accomplish, in terms of manufacturing products, and, at the highest level, independent of the hardware used to accomplish the task.
• the task is implemented or accomplished by performing actions according to the process definition.
• a task can be defined in terms of subtasks, e.g. by a user defining the task, and thereby already performing part of the planning, and/or the task be broken down into subtasks by a planning system.
• task in robot programming, sometimes the terms “task level” is used, as opposed to e.g. "motion command level”.
• task mainly to denote what a system should do, i.e. a target to be accomplished. Often, of course, a task in this sense is defined at the "task level”.
• the method comprises the step of, after determining the set of Hardware Modules,
• Hardware Modules and the compatibility of Hardware Modules with Software Modules, and generating a list of compatible combinations of Hardware Modules and Software Modules. This allows to take into account interactions and dependencies between Hardware Modules and Software Modules.
• the method comprises the step of
• the method comprises the step of, after determining the set of Hardware Modules (with or without the refining step) determining separate subsets of concrete Hardware Modules, with at least a first subset comprising concrete Hardware Modules that are available at a first geographical location, and at least a second subset comprising further concrete Hardware Modules that are available at other geographical locations.
• Such further Hardware Modules are typically described by associated Hardware Module Descriptions stored in the Inventory.
• Hardware Modules that are already present at a particular production facility (or at more than one facilities that are readily accessable) with Hardware Modules that need to be transported from other geographical locations.
• the choice of the two subsets can be optimised in view of costs of e.g. operating, transporting, renting or buying the Hardware Modules from the different sets.
• the optimisation can take the availability of the Hardware Modules in a time span for which their use is planned into account.
• the method comprises the step of updating, in the Inventory, a current status of the concrete Hardware Modules of at least the second subset in the Inventory to reflect the fact that the concrete Hardware Modules are installed and operated as part of the robotic system having been designed, and updating historical data collected in the Inventory in accordance with operating data from these concrete Hardware Modules.
• one or more of the set of Hardware Modules are predetermined as being of a particular Hardware Module type and/or as being a particular concrete Hardware Module. Such a pre-selection of types or individual modules can be done by the user, or by a design system.
• the Inventory is a unique opportunity to measure the behaviour of their Hardware in different environments and to have a true understanding of the boundaries of its Hardware. It also allows to, based on the historical data, identify aspects where improvement is required. Preventive maintenance and replacement can be scheduled, taking into account the specific conditions and configuration under which the Hardware ist used.
• each one of the performance criteria is associated with one of at least two performance classes, a higher performance class and a lower performance class;
• the performance criteria of the lower performance class are relaxed either manually or automatically until the second set of Hardware Modules and/or Software Modules that satisfy the performance criteria of the higher performance class and the lower performance class is not empty.
• Selecting the set of Hardware Modules and/or Software Modules that satisfy the performance criteria can be done with a known mathematical optimisation method.
• the underlying optimisation problem will typically be a multi-objective and multimodal problem.
• the optimisation (or input) variables comprise the choice of Hardware Modules and/or Software Modules and of their parameters.
• the performance criteria can specify fixed boundaries that cannot be crossed, and objective functions to be minimised or maximised.
• Selecting the set of Hardware Modules and/or Software Modules 4 that satisfy the performance criteria can be done, for example, with the following steps:
• the Inventory can be implemented and provide data for designing robotic systems with or without incorporating planning steps as part of the design process. Such planning steps, as described below, can be used to further aid in the selection of Hardware Modules:
• a method for operating the robotic system wherein the robotic system comprises a given set of concrete Hardware Modules and Software Modules, and wherein the location of the Hardware Modules in space (that is, within the robotic system and relative to other Hardware Modules of the robotic system) is known,
• the process definition is retrieved from an Inventory, by matching a process definition stored in the Inventory to the target task. This allows to perform planning with a planning system that is in constant communication with the Inventory, uses information from it and updating it.
• the process definition is input by a user.
• Planning can be performed in a standalone system, such as the CCC.
• the tasks and subtasks implicitly define a graph which represents mutual dependencies of actions, in particular whether actions can be performed in parallel or have to be performed in sequence. This graph can be determined explicitly, or it can be traversed without having been determined explicitly, by splitting up the subtasks recursively and performing the resulting actions as the processing of the products takes place.
• "determining a Hardware Module” involves choosing one of different production cells that are able to perform the target task.
• the method comprises the steps of
• the system can react flexibly to unexpected events. Depending on how long a action takes and on whether it is successful or fails - be it accomplished by a machine or a human worker - the system can adapt other actions and the allocation of resources, in particular of Hardware Modules, according to the current state.
• the method comprises the steps of
• Modules to actions as production progresses flexibly adapting the allocation to the evolving production process.
• the planning system can e.g. distribute the workload over other Hardware Modules or other resources, such as alternative production systems, human workers, etc. in order to maintain accuracy or to delay maintenance.
• the current status includes the location and pose of the Hardware Modules in space. This allows to coordinate movement of robots to avoid collisions.
• the Hardware Module is a smart pluggable module of a robotic system and comprises at least one sensor for measuring an internal property of the Hardware Module, a communication unit for communicating with other Hardware Modules, a data storage unit and an embedded controller,
• the embedded controller being configured to collect collected data, the collected data comprising
• the collected data is determined from sensor data from the at least one sensor
• the sensor data being transmitted can be transmitted without having been stored in the data storage unit, or it can be first stored in the data storage unit, then retrieved from storage and then transmitted.
• the internal property of the Hardware Module that is measured by the sensor is, for example, a relative position of a joint of the Hardware Module, or a position of an actuator, or a temperature inside the Hardware Module, an elongation of parts of the Hardware Module (measured e.g. by strain gauges), forces and torques acting on the Hardware Module, vibrations occurring during operation of the Hardware Module, etc.
• Some of such internal properties can be used in the Hardware Module itself for controlling operation of the Hardware Module, such as a joint position measurement being used to control a corresponding joint actuator.
• Internal properties can be used to detect malfunctions, critical conditions that need maintenance, or for adjusting calibration parameters, etc. Some internal properties can be stored and/or transmitted without being used by the embedded controller for the operation of the Hardware Module itself.
• the Hardware Module is a manipulator module comprising two mechanical links connected by a joint, an actuator for setting a position of the joint and thereby a relative position of the links.
• the robotic system comprises at least two Hardware Modules, each Hardware Module being a pluggable module and comprising at least one sensor for measuring an internal property of the Hardware Module, a communication unit for communicating with other Hardware Modules, a data storage unit and an embedded controller,
• the embedded controller being configured to collect collected data, the collected data comprising - status data representing the current status of the Hardware Module;
• the collected data is determined from sensor data from the at least one sensor
• the embedded controller being configured to perform at least one of
• the robotic system further comprising a central computation and command unit configured to
• the at least one Hardware Module is a manipulator module comprising two mechanical links connected by a joint, an actuator for setting a position of the joint and thereby a relative position of the links,
• a method for configuring the robotic system comprises the steps of
• a display device typically a video screen.
• available for use can mean that the Hardware Modules are located at a particular plant, or that they are owned and under control of the end-user's organisation.
• the list characterising available Hardware Modules can be empty, meaning that all the Hardware Modules selected by planning will have to be procured and/or transported to the location where the system shall be implemented.
• the list characterising available Hardware Modules can specify, for each available Hardware Module, its identity. Based on this, the subsequent steps can use the individual characteristics of each Hardware Module.
• the list specifies only the type of each available Hardware Module. Based on this, the subsequent steps can use the type specific characteristics of each Hardware Module. When the system is assembled, each Hardware Module can provide its individual characteristics from its internal storage. Then these individual characteristics can be used in operation of the robotic system.
• the process definition can be of the kind described earlier.
• the method is particularly suited for configuring smart pluggable modules as described herein.
• the step of automatically determining, based on their physical characteristics, a set of selected Hardware Modules comprises the steps of
• the step of automatically determining the set of Hardware Modules comprises the steps of
• the step of deteraiining additional Hardware Modules comprises the step of
• the step of determining the additional Hardware Modules comprises the steps of
• the manipulator modules are able to provide at least two types of information to the CCC unit, wherein the said two types of information comprise:
• Physical characteristics of each manipulator module and the tolerances for those physical characteristics can comprise weight, torque or force ranges, speed ranges, geometric parameters of the mechanical connection elements, and the joint position of the manipulator module or the geometric relation between mechanical interfaces, as determined by the joint position.
• the central computation and command unit controls operation of the one or more actuators of the one or more Hardware Modules by sending motion commands via communication units of the Hardware Modules.
• the central computation and command unit can use feedback from sensors of the Hardware Modules or from sensor modules.
• a manipulator module can be configured to receive motion commands from the CCC and to drive the actuator towards a joint position specified by the motion command.
• a motion command can comprises a set point for a joint position, or a trajectory comprising several joint positions, each to be reached at a specific point in time, or a trajectory of speed vs. position to be followed.
• the method for operating the robotic system comprises the steps for planning operation of the robotic system by
• the start configuration and a target configuration of the robotic system is assumed to be generated by the end-user or a higher level planning system.
• the computations required for maintaining the computational model and performing the model-based motion planning can be done on the central computation and command unit, or on one or more additional data processing units, e.g. cloud based processing units.
• the central computation and command unit receives information that defined the configuration of the robotic system or assembly from the manipulator modules and determines a mathematical or computational model representing the real robotic system, and its functional and mechanical properties. These properties can include system boundaries in terms of action range (space), payloads (weight), speed and acceleration.
• the method for operating the robotic system comprises the steps for automatically determining a computational model of the robotic system by
• an associated Hardware Module Description comprising a description of physical characteristics of the Hardware Module; o its geometric relation to one or more adjacent Hardware Modules;
• Hardware Module Description can be retrieved from a database, i.e. the Inventory, that is separate from the Hardware Module, or as stored in Hardware Module itself.
• the geometric relation of a Hardware Module to one or more adjacent Hardware Modules can be determined from the spatial relation between interfaces of the Hardware Modules. Together with joint positions of each Hardware Module, the complete configuration of the kinematic link formed by the Hardware Modules is determined.
• the physical characteristics can comprise at least parameters of a kinematic link formed by the Hardware Modules, such as Denavit-Hartenberg parameters. These can be sufficient for motion trajectory planning.
• the physical characteristics can also comprise approximate or exact 3D body models of the Hardware Modules. These can be combined to form a 3D model of the robotic system that can implement collision avoidance as part of motion trajectory planning.
• Plugging a manipulator module into a modular robot system can provide at least two types of information to the central computation and command unit (CCC), such as: The position and functions of the said modules with respect to the said assembly;
• CCC central computation and command unit
• the method for operating the robotic system comprises, for automatically determining the geometric relation of a Hardware Module to one or more adjacent Hardware Modules, the step of determining which of several possible relative spatial positions two adjacent Hardware Modules are in,
• the method for operating the robotic system comprises the following step for automatically determining the identity of one or more adjacent Hardware Modules:
• the method for operating the robotic system comprises one or more of the following methods for active tolerance compensation:
• Hardware Module This can be in the Hardware Module itself and/or in an external database. When planning and/or executing actions involving the Hardware Module, the stored measured characteristics are taken into account. This allows to produce with relatively larger tolerances and actively compensate for deviations from reference values.
• a computer program product for performing one of the methods described above is loadable into an internal memory of a digital computer or a computer system, and comprises computer-executable instructions to cause one or more processors of the computer or computer system execute the method.
• the computer program product comprises a computer readable medium having the computer-executable instructions recorded thereon.
• the computer readable medium preferably is non-transitory; that is, tangible.
• the computer program is embodied as a reproducible computer-readable signal, and thus can be transmitted in the form of such a signal. Further embodiments are evident from the dependent patent claims. Features of the method claims may be combined with features of the device claims and vice versa.
• Fig. 1 elements of a robot system
• Fig. 2 a manipulator module
• Fig. 3 a manipulator module in a different joint position
• Fig. 6 a physical structure of a robot system or assembly
• Fig. 7 an interface structure of a robot system or assembly.
• FIG. 1 schematically gives an overview of elements of a robot system, comprising Hardware Modules 3 and Software Modules 4, collectively referred to as "modules".
• Hardware Modules 3 are combined and configured to work as actuators and sensors.
• Hardware Modules 3 can be physically connected to form manipulators such as robot arms.
• Hardware Modules 3 can be entire (non- modular) manipulators or other devices such as numerically controlled machines, and sensors returning digital (on/off) values or analogue values, including cameras with or without image processing capabilities.
• Such Hardware Modules 3 can be arranged to cooperate with each other in handling real world objects.
• a translator 3a can be implemented, with an associated Hardware Module Description 5, which packages the functionality of the legacy device and makes it appear, to the robot system, like another Hardware Module 3.
• Hardware Modules 3 can be manipulator modules 33, and a set of connected manipulator modules 33, connected to a base Hardware Module 3b, forms a robotic system or robotic assembly 3c.
• Software Modules 4 reside in a distributed processing environment which implements functional entities at different levels of control and planning, such as real-time controllers for closed loop sensing and control, motion planning, collision avoidance, coordination of manipulators, production scheduling, user interfaces, calibration, communication, etc.
• these functional entities are executed in a distributed on data processing units that are realised - with regard to physical proximity and/or acceptable communication delays - closer to or farther away from the Hardware Modules 3.
• closed loop control of a Hardware Module 3 is performed in a processor of a Hardware Module 3 itself or in a processor closely associated with one or more Hardware Modules 3.
• Coordination of and planning for production cells can be performed on a supervisory computer, and overall production optimisation and planning can be performed with cloud based processing units. Together, the data processing units form a distributed processing environment 91.
• Module shall be used to refer to both Hardware and Software Modules.
• Each Hardware Module 3 is represented by a Hardware Module Description 5 which specifies, in machine readable form, capabilities and interfaces of the Hardware Module 3.
• Each Software Module 4 is represented by a Software Module Description 6 which specifies, in machine readable form, capabilities and interfaces of the Software Module 4.
• a Hardware Module 3 can be, e.g., a manipulator module 33, a base Hardware Module 3b, a central computation and command unit 10, or a sensor module 3s, or a legacy device connected an controlled through a compatibiliser unit or translator 3a..
• a manipulator module 33 in addition to having a computing unit as the other types of Hardware Modules 3, comprises an actuator (or motor) 39 and can comprise its own sensors 38, e.g. for forces and torques generate by or acting on the manipulator module, and can communicate sensor data to other Hardware Modules 3 by means of a communication unit 37.
• the computing unit of a manipulator module 33 typically is an embedded controller 35.
• a manipulator module can be physically connected, by means of one, two or more physical connections or interfaces 31 , 32, to other manipulator modules, which together form a modular manipulator system or assembly.
• a physical connection or interface 31 , 32 typically comprises a mechanical interface with mechanical connection elements for connecting the manipulator module to other manipulator modules, and an electrical interface with electrical connection elements for communication and power links.
• the manipulator module 33 is able to communicate with these other manipulator modules, to determine their identity and to exchange its identity and parameters with them and optionally with a CCC unit.
• a geometric relation between two or more mechanical interfaces 31, 32 can be set.
• such a relation can be described in terms of a joint position of the manipulator module. If the manipulator module implements a rotary joint, then the joint position is described by an angle, and the geometric relation between the mechanical interfaces can is determined by this angle and the geometric relations between the joint and each of the mechanical interfaces.
• Fig. 2 schematically shows a manipulator module 33 with an embedded controller 35 arranged to control an actuator 39, read sensor data from one or more sensors 38, store data to and retrieve data from a local data storage unit 36, and communicate through a communication unit 37.
• Main functions tasks of the embedded controller 35 can be:
• Each Hardware Module 3 knows its characteristics and is able to describe itself. Each module is characterized by at least two types of parameters amongst
• Each smart pluggable module is unique and has its own control loop
• Each Hardware Module 3 can understand and implement commands from the central computation and command unit 10, and can turn it into action.
• An action can be a movement, but can be also wait, sleep, transfer data, etc.
• Sensors 38 are driven by the embedded intelligence or embedded controller 35 of the module. Their functions can be of one of three types:
• system control to support realisation of an action or give indication on the result of the action
• Sensor readings can be transmitted to the embedded controller 35 through wire-based or wireless channels. Examples for properties measured by sensors are temperature, humidity, accelerometer, vibration, acoustical signals, etc.
• the manipulator module 33 comprises two mechanical links, a first link 31a and second link 32a, a relative position between these links being controllable through the actuator 39.
• the first link 31a comprises a first interface 31, and the second link 32a comprises a second interface 32.
• Each of the interfaces 31, 32 comprises interface elements 31b, 32b as mechanical and electrical and communication connection elements.
• the joint 34 is a rotary joint, and the first interface 31 and second interface 32 lie in planes that are at an angle of essentially 45° to the axis of rotation of the rotary joint 34. This allows to rotate the two interfaces from a position in which they are parallel to one another, as shown in Fig. 2, to a position in which they are at a right angle, as shown in Fig. 3.
• Fig. 4a shows a manipulator module 33 with a rotary joint, at different joint positions.
• Fig. 4b shows a manipulator module 33 with a joint that allows for both translation and rotation of the second interface 32 relative to the first interface 31.
• Fig. 6 schematically shows a physical structure of a robot system or assembly, with manipulator modules 33 connected to form a sequential structure, starting with a base Hardware Module 3b. In other embodiments, not shown, more than one sequence of manipulator modules 33 can be based on the same base Hardware Module 3b. In further embodiments, not shown, manipulator modules 33 or Hardware Modules 3 in general have more than two interfaces, and thus tree-like structures can be assembled with them.
• Fig. 7 schematically shows an interface structure of a robot system or assembly: power supply lines 31p and communication lines 31c run, starting from a central computation and command unit 10, sequentially through the manipulator modules 33.
• the communication lines 31c can be physically separate lines for the two directions of communication, from and to the central computation and command unit 10, or both directions of communication can pass through the same physical lines, e.g. a communication bus.
• a central computation and command unit 10 comprises data storage and processing elements and is capable of executing programs, for example Software Modules 4, for controlling and/or coordinating the movement of Hardware Modules 3, taking into account information, including sensor data, from other Hardware Modules 3, in particular from sensor modules.
• the central computation and command unit 10 controls the Hardware Modules 3 and in particular one or more manipulator arms to perform actions based on the state of the environment. This can involve
• a CCC unit can communicate with databases that are maintained by the CCC unit itself, and/or with databases maintained in computers that are external to the CCC unit, e.g. in the cloud.
• the CCC, or a Hardware Module 3 itself can store data associated with the Hardware Module 3 and its operation, and/or information gained from teaching particular tasks.
• a minimal robotic system comprises a CCC and either a manipulator subsystem or one or more manipulator modules that together form a modular manipulator system.
• the central computation and command unit 10 is drawn in a separate box, however its computational resources are part of the distributed processing environment 91. So the central computation and command unit 10 can implement at least part of the functionality of the BOS 1 and the inventory 2.
• Programming the robotic system, using the central computation and command unit 10, can be done in one or more of the following modes:
• coordinate mode the user defines coordinates of the location that the manipulator needs to reach.
• the CCC calculates the best / optimized paths to achieve the desired location and transmit corresponding motion commands or trajectories to each modules.
• Operating the robotic system, using the central computation and command unit 10, can involve one or more of the following functions:
• Action requests can come either from directly from and end-user through the user interface, or from other robots, machines, or entities through an operating system of a higher level entity, such as a group of collaborating robots, a manufacturing cell, factory, etc.
• Implementing an action can be done by iteratively planning and simulating the action on the basis of the mathematical model in order to determine the sequence of steps or movements to be implemented by each of the modules.
• Each controller reads only the actions directed to him and does not take into account the actions relevant to other modules.
• Each manipulator module 33 executes the requested actions in the sequence given by the CCC, can adapt them to sensor feedback from its own sensor 38 or from other Hardware Modules 3, and gives feedback to the CCC.
• the software modules can be integrated by a «basic operating system » 1 (BOS) which implements an interface between Hardware Modules 3, other Software Modules 4, and data storage means, such as libraries comprising software, libraries comprising hardware.
• BOS basic operating system » 1
• the basic operating system BOS 1 comprises fundamental functions which enables at least to operate a standalone robot implementing a set of basic operations.
• the BOS 1 works as an interface between the Software Modules 4, some of which communicate with and control Hardware Modules 3, thereby enabling control of a robotic system built from the Hardware Modules 3.
• a Software Module 4 can exist independently of a Hardware Module 3, for example, when it accomplishes data processing without the need for specific hardware.
• a Software Module 4 can exist in association with a Hardware Module 3, for example when execution of a function provided by the Software Module 4 involves the use of a specific Hardware Module 3.
• FIG. 5 schematically shows Hardware Modules 3, Software Modules 4 with associated Hardware Module Descriptions 5 and Software Module Descriptions 6.
• Each Hardware Module 3 is represented by a Hardware Module Description 5 which specifies, in machine readable form, capabilities and interfaces of the Hardware Module 3.
• the Hardware Module Description 5 can comprise
• hardware ID 51 a unique identifier of the Hardware Module 3;
• hardware parameters 52 parameters describing properties of the Hardware Module 3 that to not change, for example, a nominal length and weight of a robot arm, the size of a fixture.
• the hardware parameters 52 can include a 3D geometric model, allowing to plan for collision avoidance when planning robot movement;
• hardware location 53 the position and orientation of the Hardware Module 3, typically within a universal production cell;
• hardware state 54 parameters describing an internal state of the Hardware Module 3, for example, a current position of an actuator;
• calibration data 55 typically properties of the Hardware that change slowly and are measured from time to time, such as deviations from nominal hardware parameters 52;
• • history 56 for example, number of operations performed, parameters of these operations, number of successful and unsuccessful operations, related error messages, etc.
• Each Software Module 4 is represented by a Software Module Description 6 which specifies, in machine readable form, capabilities and interfaces of the Software Module 4.
• the Software Module Description 6 can comprise
• module ID 61 a unique identifier of the Software Module 4;
• module method 62 one or more methods implemented by the Software Module 4;
• module state 63 an internal state of the Software Module 4.
• module resource 64 resources required by the Software Module 4. This can be Hardware Modules 3 that the Software Modules 4 is configured to control. This can also be computing resources that the Software Module 4 requires to run itself.
• the Hardware Module Descriptions 5 and Software Module Descriptions 6 are maintained, in the distributed processing environment, in an Inventory 2.
• Hardware Module 3 When a Hardware Module 3 is connected to a system, and in particular when a manipulator module 33 is connected to a robotic system, its presence is detected, e.g. by a central computation and command unit 10 to which it is connected directly or through other Hardware Modules 3.
• the BOS receives information that the Hardware Modules 3 provide about themselves, such as physical characteristics of manipulator modules, and their position and function within a modular manipulator system. From this information, the BOS creates the virtual infrastructure, that is, a virtual representation of the modular manipulator system and its functional and mechanical possibilities, and further infrastructure such as other manipulators, hardware, production line, .... as represented by associated Hardware Modules 3.
• the BOS can simulate within the virtual representation actions that are required from the CCC, and then
• This virtual representation can also be used in a configuration tool for sales support.
• the execution system to implement a required action or command, the execution system:
• o Optimise according to predefined criteria This can be done by iteratively or in parallel simulating and evaluating a plurality of different implementations of the command or action. o Make workload repartition. This can be done by identifying time critical processes, tasks or subtasks and distributing the workload better or propose additional hardware and/or software modules, o Check the advantages / disadvantages of new Apps
• the Inventory 2 serves as a library registering information in relationship with Hardware Modules 3 and Software Modules 4, and in particular Hardware Module Descriptions 5 and Software Module Descriptions 6.
• the Inventory 2 exchanges information with the Basic Operating System 1.
• the Basic Operating System 1 can retrieve information about the physical characteristics, current state and historical data regarding a module, and also regarding compatibility between Hardware Modules 3 and Software Modules 4.
• Storing a module's history in the Inventory 2 allows to give information about the module's status and history. The status can, if a module is not in operation, include the module' location in a hardware warehouse.
• Information regarding a module can be generated by the module itself and transmitted to the Inventory 2 directly and/or via the BOS. For this, a standardised structure of the information and a protocol for transmitting the information is used.
• an interface unit or compatibiliser unit is added, configured to provide on one side an identity of the device and the standardised information and communication means, and on the other side means for retrieving information from the third party device.
• the Inventory 2 is monitored in order to detect updates to Software Modules 4 (and also Hardware Modules 3). If an update is detected, or if a function is detected that promises to perform a particular operation in a better way (perhaps with drawbacks in another area), this information can be transmitted to Hardware Modules 3 or entire installations that can make use of this operation. Similarity of operations can be determined on the basis of their descriptions. Heuristic learning methods, e.g. using neural networks.
• This section gives an example for planning the execution of a task in a concrete manufacturing system, which can be a robotic system or a production cell.
• the example can be adapted to the planning of a task in a generic production cell, in which the hardware equipment is simulated.
• the example can also be adapted to re- planning, where the planning starts at some intermediate state of the production cell - be it concrete or simulated - in which some production steps have already taken place, instead of an initial state in which no production steps have taken place yet.
• a pre-requisite is that the production cell has been designed and is defined in terms of:
• Hardware Modules present
• the BOS is connected to the inventory as well as to the concrete Hardware. Specific software modules which have been selected during the design phase have been installed either on the Hardware equipment or on the BOS.
• the BOS is aware of the available HW equipment, of their status in real time, and of the available software modules.
• the inventory gives the BOS access to Hardware Module Descriptions 5, such as Hardware specifications, historical data related to the Hardware, to process definitions and to software at a general level.
• Hardware Module Descriptions 5 such as Hardware specifications, historical data related to the Hardware, to process definitions and to software at a general level. This can comprise information as in the following table:
• the inventory does not have the information for each specific, concrete task definition for example, but rather for types of tasks on an abstract level.
• a company can thereby provide abstract task know-how without disclosing special process know-how. This allows protecting confidential process data from a third party.
• Planning of a task starts with defining an Entry Status E and an Output Status O of the task.
• these describe the status or state of a product that is to be manufactured.
• they can also describe the status or state of the manufacturing system.
• the task can either be mapped to a set of actions, or it is split into subtasks, which in turn is mapped to actions or recursively split into further subtasks. This can be done either by a human or by the BOS.
• Subtasks are also defined by entry and output statuses.
• the BOS 1 maintains a computer representation of the available production cells as well as the equipment, that is, the Hardware Modules 3, available within the cells. It queries the inventory to compare the task to be implemented with the task definition stored. It can then find out some similar tasks and their process definition, and a list of required equipment or Hardware Modules 3 for accomplishing the task. Similar tasks can be tasks that have the same structure but under different conditions and environments. For example, picking and placing an object or drilling a hole. Picking and placing a computer screen is similar to but not the same as picking and placing a steel ball. The BOS can then compare this list with the concrete equipment that is available within the production cells. It can then identify the production cell which best matches the list of "must have" required equipment. If necessary, it can generate a list of Hardware Modules 3 that are required but are not available.
• the task is then split in subtasks by the BOS. Those subtasks can be realised with the available equipment within the production cell and enabling to achieve the Output O of the task. If necessary, the BOS 1 can generate suggestions for additional non- available equipment.
• Each subtask is defined by an entry status and an output status. These status described the state of the part to be processed in terms of visible or measurable characteristics, for example :
• Entry status and output status usually are different. Typically, only a subset of such characteristics and their change can be specified for each task. The other characteristics are unaffected by the task, or are not of interest.
• Sensors for contact or noncontact measurements are able to detect or measure the change of status between entry and output.
• the sensors are connected to the BOS and able to transfer the data to the BOS.
• the BOS can perform a comparison between the measured status and a status according to the process definition to know at which stage of the process the part is. This can be done to check the status, to see whether the execution of the process definition is going according to plan. It can also be done to re-plan execution if a part enters the production cell unexpectedly, i.e. not under control of the BOS 1, or after a failure of an action.
• entry and output status such as x, y, z express that one or more characteristics have certain values. More concrete examples for this are given in the subsequent section on planning.
• A4 Since the entry status of A4 is the output status of A3, it follows that A3 must be executed before A4.
• One or more of the subtasks could be actions, i.e. not subject to being split up into further subtasks.
• the process definition comprises subtasks B and B2 which can be er ormed inde endentl :
• the entry status of B3 requires that Bl and B2 be performed. But the entry status of Bl and B2 are unrelated, so Bl and B2 can be performed in any order or simultaneously.
• the Process definition comprises subtasks CI and C2 that could be swapped:
• the table entries signify, for example, that Subtask CI can change the status from E to xl , or it can change the status from yl to x2 (but not from E to x2 or from yl to xl). Consequently, in order to get from entry status E to output status O, there are two paths:
• the BOS While performing actions for accomplishing one or more tasks, the BOS knows the occupation of each Hardware Module 3, or generally of resources within the production cell in real time. It also knows the status of the parts being processed. This can be accomplished thanks to continuous or frequent observation of the production cell and the parts by the sensors. It allows to adapt the planning, or to re- plan, depending on the outcome of actions being performed. When doing this, it is not necessary to plan all future actions specified by a task definition and its recursive reduction to subtasks and finally actions. Another way to implement the planning process is to determine the mutual dependencies of subtasks and finally actions, and to civilistically perform whatever actions can be performed according to the task definition and for which resources and in particular Hardware Modules 3 for performing the actions are available.
• exemplary concrete status are:
• the production cell detects the status thanks to a camera equipped with an app for color detection and a balance for weighing.
• the camera and balance are examples for Hardware Modules 3 that are sensor modules.
• requirements can specify the need for two robots, or of a human and a robot.
• the BOS will then select the adequate Hardware Modules 3 available to perform the task in collaboration with other Hardware Modules 3 or a person.
• the BOS determines the actions to be accomplished by each Hardware Module 3, and their mutual dependency, e.g. as a sequence and/or by parallel execution.
• the vision system can transfer to the BOS information related to the execution of the tasks by the robots in real time, so that the BOS can cause the next order to be executed once the previous one is finished.
• Example of task definition Drilling a hole of diameter 1.2 mm into a piece of metal.
• Robot Rl is equipped with a gripper and robot R2 is equipped with a drilling tool.
• the BOS will request status from robot R2.
• Robot R2 is not yet equipped with the right type of drill bit for drilling a hole of 1.2mm into a piece of metal.
• the BOS identifies the right drill bit required thanks to information from the inventory.
• the BOS knows from the inventory that the production cell already has the right type of drill bit stored in its buffer or storage area. The exact location of the drill bit within the storage area is determined thanks to a camera assisted with vision system software.
• Robot Rl receives the order from the BOS to replace the drill bit of robot R2 with the one required for the action. This task is divided into subtasks:
• the BOS can alert a human worker and instruct him to change the bit.
• the vision system will give information about position and orientation of objects to the BOS. Once the robots are ready, the BOS will order to Robot Rl to pick and place the piece of metal at a predetermined place in a specified orientation, and to hold it. Then, once this subtask is completed, i.e. the objects are located at the right position, the BOS validates that all "must" conditions that make the "drilling" action possible are met, the BOS will order to Robot R2 to drill the part at a predetermined position.
• robots can be compliant robots that satisfy requirements for human-machine collaboration, e.g. as in Specification ISO/TS 15066 «Robots and robotic devices— Collaborative robots».
• the BOS is connected to the Inventory 2.
• the Inventory 2 comprises specifications from suppliers, general historical data and (abstract) process definitions.
• the task is part of performance requirements, and Hardware Modules 3 and their configuration are determined such that the robotic system satisfies the performance requirements.
• An example for this process is given in the following, first in general and then for a concrete case.
• a task definition can define what needs to be performed, how many times, with which quality grade, which accuracy. What needs to be performed can be defined in terms of outcomes and/or in terms of actions.
• a task definition can be specified by one of various known programming paradigms or by a combination thereof: GUI manipulation of a computer model, teaching mode with a concrete manipulator, parameterised motion commands, learning by demonstration, etc. ... based on data entered by a human operator, from CAD files, from a vision system, from force and other sensors, etc.
• the BOS with an appropriate software, transcribes these observations of movement, trajectories, orientation, etc into actions to be executed by one or more Hardware Modules 3.
• the human operators can add required tolerances, the number of parts, their weight and all specific requirements in terms of performance criteria.
• the BOS Given a target task, the BOS is able to find within the inventory 2 comparable tasks and to divide them into subtasks. For this, the target task definition is compared by the BOS with a list of abstract process definitions stored within the inventory. The following algorithm can then be implemented:
• the BOS can propose, thanks to the inventory 2, a list of equipment and in particular Hardware Modules 3, according to predefined or user-defined criteria such as: price, energy consumption, overall price including maintenance costs, rent of equipment, etc... (where the criteria can have different weights).
• the BOS may need to search within the inventory, at a first level, for types of equipment that are required. On a second level, it can search for types of components that can be assembled to build one piece of equipment. This can be, in the case of there being pluggable modules that can be assembled, or in the case of effectors and robots being combined, or in the case of machines requiring supplies of specific consumables.
• the BOS can simulate both scenarios and select the optimal scenario according to predefined or user-defined criteria.
• Such predefined or user-defined criteria could be: lowest investment, fastest production cell set-up, re-use as much as possible existing assets of a production cell or factory.
• the BOS can order assets that already are in-house to be transported to the location of the production cell. Either a remote-controlled trolley can pick them up, or a human worker, or, if it is mobile, equipment can transport itself.
• the Inventory 2 can comprise information at two levels: - describing Types of hardware from different brands with specifications and tolerances given by suppliers.
• the BOS can comprise information
• step 1 the following describes process definitions, and their relation to product design:
• the BOS Given a process Macro-definition and knowledge of the Task Output characteristics or status, and performance criteria, the BOS can access to the Inventory, meaning that the BOS:
• the selection of the process Micro definition can be given to a human operator, or the system can select the most appropriate process Micro-definition by combining and weighing the user-defined criteria: the time for setting up the installation and the cost per unit can have a weight 2, and other criteria have a weight of 1.
• the system can output a set of different process definitions from which the user can choose.
• Design a production cell whose task is: Assembly of two parts by screwing.
• the plant or a production cell already exists and comprises some assets, such as a manipulator, cameras, process table (working area, optionally with fixtures), etc.
• the process definition is determined by: Teaching mode: a person shows through the vision system the tasks which need to be performed.
• An ERP system can provide further information on parts, orders, batch sizes, etc. Parameters are collected thanks to different means like for example:
• Weight of parts (information given by human operators ).
• Shape of parts (information given by CAD).
• a reference point can be used to make the calculations for the set-up.
• This reference point refers to the location of the process table n°l in our example. The locations within the production cell are defined relative to this reference point. If no reference point is given by a human operator, the system automatically chooses process table n°l , and the reference point is placed in the center of the production cell.
• the number of preventive maintenance events should be as low as possible.
• the BOS determines a final cost per unit produced, including amortization costs.
• the BOS can initially be constrained to look at a limited set of characteristics of the design that allow to achieve the desired performance and to match the performance criteria, or begin with standardised solutions and move to specialized solutions.
• the BOS can perform the search, for example, until it has the desired number of proposals (for example 5) or until a total cost is optimized.
• the BOS can search to minimize the space required for performing the task inside a production cell, or for a production cell as a whole.
• the BOS searches the Inventory for a corresponding Process definition, for example: "screw two parts when parts are threaded".
• the process definition may be refined as follows, based on information from the Inventory:
• the BOS can first compare the process definition which the existing production cells described in the Inventory and implementing the same task. If there is such a match, the BOS is able to :
• the task is then split into subtasks by the BOS according to the process definition found in the Inventory 2.
• Subtask 8 Pick and Bring Assembly on
• Subtasks 1 , 2 and 3 can be performed independently, or simultaneously.
• Subtask 4 can be done directly after subtask 1 or after subtask 2 or after subtask 3.
• Subtask 5 needs to be performed after subtask 4 but can be performed before subtask 3.
• the actual order of the tasks when implementing the actions of the process definition is determined during planning, typically in an opportunistic manner.
• this step may be skipped.
• Subtask 1 Pick Place part Manipulator SW 1 SW 101 Manipulator + parts A, A on the G ripper SW 2 SW 1 and/OR in the process Process Table SW 3 SW3 buffer table PT 1 Buffer area Gripper + area 1 Vision System SW2
• Subtask 2 Pick Place part Manipulator SW 1 SW 101 Manipulator + parts B, B on the Gripper SW 2 SW1 and/OR in the process Process Table SW 3 SW3 buffer table PT1 Buffer area Gripper + area 2 Vision System SW2
• Subtask 3 Pick Place part Manipulator SW 1 SW 101 Manipulator + parts C, C on the Gripper SW 2 SW1 and OR in the process Process table SW 3 SW3 buffer table PT 1 Buffer area Gripper + area 3 Vision System SW2
• A+B+C area 4 Vision System The Software Modules 4 listed (SW 1, 2, 7) typically are, depending on the associated Hardware Modules 3, drivers, control algorithms, signal processing, video processing, etc. At this stage, the BOS will determine the minimal number of components required in the Production cell:
• the process table characteristics can be determined from the subtask descriptions in the following way: requirements on the characteristics are calculated from each subtask.
• the BOS compiles the requirements for all subtasks in which the process table is involved and determines the characteristics that satisfy all requirements.
• Subtask 1 the size of the part A will involve a process table minimum size of 300*400 mm, and the minimum load to be supported by the process table is
• Subtask 2 the size of part B implies a minimum process table size of 30*30mm, and a minimum load of 250g.
• Subtask 4 the minimum process table size is 400*400mm as it must accomodate part A and part B side by side. The minimum load to be supported by the process table is then 1030g.
• Subtask 5-6-7 the minimum process table size is 400*400mm as it must accomodate part A and part B and part C side by side. The minimum load to be supported by the process table is then 1035g.
• the BOS will take into account the highest load requirement (1035g) and the largest size required (400*400mm), and a minimum height of 900mm.
• the process table 1 PTl must be qualified for work in a clean room.
• the BOS will search for a process table with those characteristics in the Inventory, and will return, for example, the 20 closest results matching those requirements.
• the BOS defines an arbitrary buffer area size for a reasonable number of parts, considering that the buffer areas can be fed during the task without interruption of the process.
• the size of buffer areas may vary from one part to another. Nevertheless for calculations of the manipulator size or reach, the size and location of the largest buffer area is considered. For example, 25 parts A stacked together: require a volume of 180*320*1 125mm 3 .
• the BOS calculates and optimizes the space required for the buffer area.
• the manipulator type is defined according to the characteristics given.
• the BOS detects that subtasks 1, 2, 3, 4, 5 and 6 requires the same type of Hardware Module 3. Accuracy required to perform subtask 6 is nevertheless much higher than for other tasks and within the Inventory 2, it is recorded as being performed by human workers in 62% of the cases known to the Inventory 2. As a result, the BOS proposes two alternatives:
• the BOS can determine the technical requirements for the manipulator: motors, torques, stiffness, material, etc. When doing so, it can determine the minimum requirements in terms of space, force, movements, etc.
• the BOS calculates from the data given that if the buffer areas are at the same height as the process table, the manipulator needs 1 less degree of freedom.
• the BOS will look for a static manipulator having 3 degrees of freedom with a minimal arm extension of 800mm to cover the process table 1 size with a buffer area size covering at least the size of the larger part (part A).
• the tolerances for each subtask are considered.
• the BOS can: first evaluate manipulators matching the minimum requirements, and look for additional software to upgrade the manipulators to the higher upset requirements.
• the BOS will return the 20 closest results matching those requirements.
• the motion paths of the manipulator are not defined in the design phase, as they are determined and adapted on the go during planning and execution.
• the characteristics of the parts and of the movements are integrated into calculations by the BOS to determine the required force, size, material, etc.
• the BOS will then determine from the Inventory the 20 most relevant results.
• the BOS can display this list of relevant grippers and also compatible manipulators and Software Modules 4.
• the BOS knows from the inventory that this tasks is in 62% of the cases done by human workers. The BOS will first search solutions in which human workers complete the subtask. It will secondly propose alternatives with effectors, in case human workers are not available.
• the BOS can compare the compatibility between the Hardware Modules 3 themselves and between Hardware Modules 3 and Software Modules 4, and determine a list of possible combinations, as well as a list of "compatibilisers".
• Compatibilisers are Hardware Modules 3 or Software Modules 4 that enable the compatibility of two elements which are initially not compatible. For example, they are manufactured by competitors.
• the BOS can also find in the Inventory 2 existing production cells which match most of the most important criteria and that have open capacities. Based on this, it provides the possibility to externalize and subcontract production.
• the final selection involves a selection according to user-defined criteria, such as:
• the BOS will compare each solution to the list of existing assets already in-house, their availability and their maintenance schedule.
• the solution chosen will give priority to the in-house assets.
• the BOS can also optimise the overall timeframe and costs associated to the completion of a production cell, or provide the information that allows the user to make a choice.
• Solution B and renting hardware and software may be the most efficient solution.
• the required elements are either bought or rented. In other cases, some may be bought and some may be rented to optimise the time to start production and/or the overall costs. Please note, that this comparison to "assets in-house” can also be done at step 4, and proposed solutions built on the existing assets.
• the BOS In order to support the realisation of the production cell, the BOS generates a mapping of the integration of the different elements, that is, their physical layout and configuration, and of infrastructure requirements like power source, light specifications, etc.
• the positioning of the equipment is based on the position of the reference point which is here the position of the middle of the Process Table 1.
• the BOS also can generate a list of orders to be placed, and book the required existing assets in the system for the planned production period. If the solution of subcontracting is chosen, this step can be skipped:
• the BOS may issue an order to the ERP for the reservation of the installation in Spain.
• the new production cell is connected and registered in the Inventory 2 for collecting data generated during production. Then, by calculating performance and comparing it to data available in the Inventory 2, the BOS can propose optimizations of the process. New pieces of software or new hardware used in a specific configuration can be added to the Inventory 2.
• the size and capabilities of the production cell are not necessarily limited to one task. It has been simplified for the sake of clear explanations.
• the design method can be applied to a single production cell or to an entire plant, in each case for just one task or for a collection of tasks. Designing or re-designing a plant or production cell, including the implementation of modifications, can be done while they are running. Likewise, planning and implementing plans can be done while the system is running, with the implementation of plans opportunistically using hardware and other resources when they become available.
• the Inventory 2 can also store configurations of production cells corresponding to process definition. Based on this, to design a new production cell, the BOS can search the Inventory for a design of a production cell corresponding to the process definition. Then the BOS can make adjustments, thereby copying and modifying the production cell set-up.
• the target of the system is not primarily related to speed of production or production cost per unit. Rather, a target is to produce as early as possible, even in a suboptimal mode.
• the end-user defines critical criteria in terms of a acceptable ranges which need to be matched during production. For example, in step 4 of the design method (determining lists of possible hardware and software) the criteria can be classified according to their importance or priority:
• Class A must have - first priority
• Class B must have - second priority
• Class C "nice to have” - third priority
• the BOS will search within its initial list of potential Hardware Modules 3 and Software Modules 4 for those that match the class A criterion. From the resulting subset, a reduced subset is generated, comprising the modules that also satisfy the class B criterion. Class C and D criteria may not be fulfilled in a first iteration. If this is the case, then depending on user input, the BOS can search for modules that are not in-house, or suggest relaxing the requirements. This can be done by relaxing requirements on e.g. production speed, cost, etc.
• a subtask requires a manipulator with the following specifications:
• the priority of Accuracy is a class A criterion; whereas, speed is a class B criterion, and Weight is a class C criterion.
• the BOS will start the search for manipulators with respect to class A. Then it will select amongst the subset of manipulators matching the accuracy requirements those that match the speed requirements (class B criterion). From the resulting reduced subset of Manipulators matching the class A and B criteria, the BOS selects those which matches also criterion C: weight requirements.
• BOS does not find any results matching criteria A AND B AND C
• BOS can propose some compromises on criterion C: for example, the tolerance for the maximum weight is extended to 5%. And then BOS can propose relevant solutions that will satisfy the most important requirements.
• Parameters of the Hardware Modules 3 can be determined by estimation or measurement during operation, and/or in dedicated calibration procedures. At some time, it may be found that achieving an adequate accuracy is no longer possible. This can involve the following steps:
• Determining if the parameter which does not fulfil requirements is critical for example, based on the priority class of a criterion related to the parameter. If it is a critical parameter, a then search within the Inventory 2 for a set of Software Modules 4 able to provide correction or compensation of errors. Within the set, the Software Module 4 is selected which allows correction or compensation of errors caused by or related to said parameter.
• the BOS indicates the need for improvement of the software or the missing software type to the End-user or to a community programming platform.
• This number can be determined, for example, from a number of units to be produced in a production run for which the design is intended, and from a number of cycles required for each unit. The latter can be determined by simulating operation of the Hardware Modules 3 for production.
• the BOS can choose a combination of Hardware Modules 3 that allows to perform the complete production run without maintenance, or with maintenance times of multiple Hardware Modules 3 being at the same time, minimising overall downtime, or by distributing production over several production cells, or by choosing Hardware Modules 3 that requires less maintenance.
• One or more or all steps or selecting Hardware Modules 3 can be performed by expressing the selecting problem as a multivariate optimisation problem, wherein parameters to be varied are one or more of
• Historical data of hardware modules 3 are collected in the inventory. Using this data, for each Hardware Module 3 one or more of the following parameters can be determined, and also statistical values derived from these parameters:
• the parameters and associated statistical data are also considered to be historical data.
• the Inventory can collect the data not only individually, associated with each unique piece Hardware Module 3, but also for each Hardware Module 3 type.
• Such Hardware Module 3 type data can comprise averages and other statistical parameters that characterise parameters or historical data collected from a plurality of Hardware Modules 3 of the same type.
• the BOS can choose Hardware Modules 3 according to different user-defined criteria: immediate availability, costs, etc. Thanks to the inventory, it is be possible to have a more global and integrated overview of the real costs of operation, including maintenance costs, energy consumption, downtime, etc. and the manner in which the costs depend on the choice of individual Hardware Modules 3.
• the BOS can output, in the step of "final selection of the Production Cell", the list of required Hardware Modules 3 together with their overall costs per module, and optionally also a correlation between the choice of a Hardware Module 3 and the total cost of operating the entire production cell, helping the end- user to make an optimal choice.
• Hardware Module 3 module and/or from data of others of the same type
• the Inventory 2 can collect historical data related to Hardware
• Module 3 types With regard to maintenance, this can include maintenance periods, parts involved, number of cycles performed between maintenance events, cause of maintenance, data describing repairs, etc. Corresponding statistical data (computed by the BOS or by the Inventory 2) per type of HW can be used to determine adequate intervals, e.g. in terms of time or cycles, for predictive maintenance, as well as tasks which need to be performed for maintenance.
• the BOS has access to the number of cycles performed by each individual Hardware Module 3 - characterised by a unique hardware identifier (ID) - as well as the number of cycles which still needs to be performed.
• ID hardware identifier
• the number of cycles performed in comparable conditions can be compared to the average number of cycles performed by the same type of Hardware Module 3 between two maintenance events. From this, an expected number of remaining cycles per Hardware Module 3 until the next maintenance event can be determined.
• the BOS can anticipate and schedule transferring production to another production cell when scheduling the next planned maintenance, or coordinate maintenance operations for all modules involved in production to take place at the same time.
• the inventory 2 can also collect and associate historical data in association with a particular configuration of Hardware Modules 3, for example, for a robotic assembly 3c comprising several manipulator modules 33.
• Such robotic assemblies can be considered to be of the same type if they are made of Hardware Modules 3 of the same types in the same configuration.
• the BOS can collect, analyse and compare, in association with conditions of use - such as the type of tasks - historical data from all concrete instances of this type. It can then estimate when failure of the robotic assembly 3c can occur: for example, in a particular configuration, joint No. 2 fails after 9746 movements handling an average load of 21 kg.
• the BOS can plan the maintenance of the equipment. In the case of planning, the BOS may choose to use a robot of a particular type only if the remaining estimated number of cycles to be performed is high enough to perform the task or subtask required for the whole production run.
• the interface to the BOS is known and different manufacturers have their own interfaces.
• An integration provider or a user community or hardware manufacturers can develop Software Modules 4 that makes the Hardware Modules 3 compatible for communication with the BOS.
• Fig. 8 shows the structure of a standalone or autonomous system 23, comprising Hardware Modules 3 with Hardware Module Descriptions 5, wherein the Hardware Modules 3 can be sensor modules, manipulator modules 33 etc. as already described earlier. These form pluggable modules with embedded intelligence and sensors. They have the ability to collect and store their own data.
• the autonomous system 23 also comprises a central computation and command unit 10 (CCC).
• the CCC is configured to perform the following functions
• Inventory 2 to receive updates, and - depending on user preferences and contractual agreements with other parties - upload and download historical data, and data derived therefrom, in particular current data related to the planning of predictive maintenance.
• the CCC 10 can implement functions for predictive maintenance and/or for configuration of an installation in view of specific tasks.
• temperature, humidity, radiation, etc. can be used, for example. for quality control or for adapting operation of the system to environmental conditions.
• the CCC 10 Based on current updates from the external repository 22, the CCC 10 knows how the pluggable hardware modules 3 can be used or have to be maintained. Data can be sent via the CCC to the external repository 22 for extended services, e.g. to improve the collective knowledge base, however this can be optional to ensure data security.
• Determining an optimized installation for a specific activity The following functionality can be part of an online tool and store in which an end- user can configure an autonomous robot system 23 with Hardware Modules 3 that are available in-house, optionally and if necessary involving additional Hardware Modules 3. The following steps can be executed:
• the tool can still be made available for use, but with limited functionalities or with restrictions.
• the end-user may not have their existing pluggable modules loaded into the system. Then the system will not build a solution using on the end-user's existing assets.
• the CCC processes the above inputs in relation with its data and in combination with an occasional connection to the external repository 22 to propose a set of the required Hardware, and its configuration, enabling optimization of a user- defined criterion.
• the process can be as in the process described with regard the design of a production cell in general, but with different constraints: one goal is to use existing Hardware Modules 3 (in-house or belonging to the end- user's company), and the Hardware Modules 3 are mainly pluggable manipulator modules 33 which allow to assemble robotic assemblies 3c in different configurations.
• Simulating the targeted task according to the process definition and the other inputs The end-user can use the results of the simulation to validate the configuration and, if necessary, modify the input. This leads to a modified configuration, and the process is repeated. Iteratively modifying the configuration and simulating can be repeated automatically in order to find an optimal solution.
• Simulation of the final solution Again, the end-user can use this to validate the configuration and, if necessary, modify the input.
• the result of the configuration process is the physical configuration of the Hardware Modules 3 and - if necessary - a set of further Hardware Modules 3 that need to be acquired.
• a manipulator shall be used for transporting an assembly from point 1 to point 2.
• the distance between point 1 and 2 may vary from 2m to 4m. It follows that there is no need for a mobile robot.
• the assembly is a part made of wood panel measuring 300x300mm and
• the manipulator is the reference point of a Cartesian coordinate system and is set up in the middle between point 1 and 2.
• the manipulator is made of a basis, labelled module M0 - which could also be on wheels - and the additional modules are labelled Ml, M2, M3, M4..., the number relating to their position after M0 when assembled to form a manipulator module 33.
• the last module is the effector.
• the configuration or design tool refers to its database, and determines
• the design tool need not display to the end-user the level of information described in the 4th column.
• the design tool utilises a connection to the external repository 22 to have access to the database of possible modules to be acquired. It will then issue a first list of possible modules.
• the CCC can then simulate operation of the robot assembly with the selected modules and check that when simulating the realisation of subtasks, the simulated performance matches requirements. Then the CCC, automatically or under control by the end-user, could add or remove one degree of freedom, perform a second iteration of simulation, and compare the results.
• the above configuration functionality of the CCC connects to an external repository 22 in order to retrieve information on further Hardware Modules 3 that are not available to the end-user in-house. If it is determined that such further Hardware Modules 3 are needed, the external repository 22 can provide functions of a shop or web shop, allowing the end-user or the CCC to order delivery of such further Hardware Modules 3.
• the CCC and/or the external repository 22 can be implemented by means of web platforms.

Landscapes

• Engineering & Computer Science (AREA)
• Business, Economics & Management (AREA)
• Human Resources & Organizations (AREA)
• Robotics (AREA)
• Mechanical Engineering (AREA)
• Strategic Management (AREA)
• Entrepreneurship & Innovation (AREA)
• Economics (AREA)
• Physics & Mathematics (AREA)
• Automation & Control Theory (AREA)
• Marketing (AREA)
• General Physics & Mathematics (AREA)
• Development Economics (AREA)
• Theoretical Computer Science (AREA)
• Educational Administration (AREA)
• General Business, Economics & Management (AREA)
• Game Theory and Decision Science (AREA)
• Tourism & Hospitality (AREA)
• Quality & Reliability (AREA)
• Operations Research (AREA)
• Artificial Intelligence (AREA)
• Evolutionary Computation (AREA)
• Fuzzy Systems (AREA)
• Software Systems (AREA)
• Mathematical Physics (AREA)
• Manipulator (AREA)
• General Factory Administration (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Related Child Applications (2)

Publications (1)

Family

ID=60190760

Family Applications (1)

Country Status (5)

Cited By (1)

Families Citing this family (12)

Citations (12)

Family Cites Families (13)

• 2017
  • 2017-10-27 EP EP17198994.0A patent/EP3476545A1/en not_active Withdrawn
• 2018
  • 2018-10-25 US US16/759,613 patent/US20210178575A1/en not_active Abandoned
  • 2018-10-25 CN CN201880071564.0A patent/CN111386178B/zh active Active
  • 2018-10-25 KR KR1020207012097A patent/KR102567103B1/ko active IP Right Grant
  • 2018-10-25 WO PCT/EP2018/079324 patent/WO2019081661A1/en unknown
  • 2018-10-25 EP EP18789184.1A patent/EP3691835A1/en active Pending
• 2022
  • 2022-09-06 US US17/903,213 patent/US20230070428A1/en active Pending

Patent Citations (12)

Also Published As

Similar Documents

Legal Events